Phonological disorders (PD) are one of the most common types of communication disorder diagnosed in children. They have been characterized as a deficit in the misarticulating child's phonological knowledge. In theory, such a deficit could affect any or all of three distinct representational levels: (1) the acousticperceptual level, (2) the articulatory level, and (3) an intervening level of symbolic categories (phonemes) that co-index acoustic patterns with the articulatory configurations that might reproduce them. Language development involves a progressive mastery of phonological knowledge at all of these levels that continues well beyond the age when the majority of children with typical development (TD) can produce most of the phonemes of English. Previous research shows that performance of children with PD resembles that of younger TD children on tasks measuring perception and articulation skills, but there is no difference to TD age peers on a task probing the robustness of symbolic representations. These results can be interpreted as support for a more specific characterization of most PD as a deficit or delay in one or both of the modality specific representational levels, but not in the ability to form symbolic phonological categories. If correct, this interpretation suggests that therapy for PD should focus on the two modality-specific representational levels. To rule out an alternative interpretation, however, typical phonological development must be compared across languages. The task probing the development of phonemes measured repetition accuracy for phoneme sequences that are attested in few or no words of English. Older TD children may be no more accurate than age peers with PD only because some infrequent sequences have inherently more difficult phonetics. The proposed research uses nonword rePetition and picture-naming to compare phonological development in 2-6 year olds acquiring English, Cantonese, Greek, and Japanese. These languages differ in the relative number of words containing various difficult sound sequences. Three specific questions will be addressed. First, are all difficult sound sequences acquired equally late, or do young children make fewer errors on sequences that occur in many words of the ambient language? Second, when a child misarticulates a difficult sound sequence, does the outcome reflect only phonetic constraints, or are more frequent sequences substituted for less frequent ones? Third, how is accuracy affected by cross-language differences in phonetic detail?